Walking the Perilous Line Between Research and Politics
Despite six months of unrelenting criticism, including three days of private grillings in late July by some of the nation's top statisticians and education researchers, sociologist James S. Coleman says he stands by his contention that in general private high schools do a better job of educating students regardless of their backgrounds than do public schools.
In an August memorandum to the National Center for Education Statistics (N.C. E.S.), the sponsor of his controversial report, Public and Private Schools, the University of Chicago professor reportedly agreed to some technical revisions for the final version of the report.
But, N.C.E.S. sources say, Mr. Coleman's memo makes it clear that he has not been dissuaded from his overall conclusion , even though other social scientists say methodological flaws cast serious doubt on their validity.
The report has been dissected by members of the American Educational Research Association, other researchers convened by the National Institute of Education, statisticians assembled by the prestigious National Academy of Sciences, and dozens of social scientists working on their own. The next round will be a "symposium"-a series of scholarly papers and essays-in the November issue of the Harvard Educational Review.
Mr. Coleman, who has been a principal figure in some of the major education controversies of the last 15 years, says he unwittingly took a first step back into that role in 1979, when he agreed to participate in the N.C.E.S. study High School and Beyond, the largest government effort ever undertaken to gather information about American secondary schools.
Mr. Coleman's National Opinion Research Center (N.O.R.C ) was hired by N.C. E..S. to develop the overall design of the survey and to develop a profile of private schools. N.O.R.C. administered cognitive and attitudinal surveys to some 58,000 students in grades 10 and 12 in 1,016 public and private high schools across the country, and surveyed attitudes of parents and teachers as well. The resulting information, which researchers call the richest data base ever constructed in American education, was eagerly awaited by scholars and policymakers.
But Mr. Coleman, say his critics, overstepped the bounds of his research role for N.C.E.S. when he finished collecting-and began interpreting-the statistics.
A Tense Presentation
In a tense public presentation of the overall N.C.E.S. data held in April by the agency, Mr. Coleman suggested that the schools-public and private-whose students show the highest achievement levels share certain crucial characteristics: low rates of absenteeism; emphasis on academically challenging courses; frequent assignment of homework; and orderliness.
And, he said, private schools as a group perform better on all counts.
Predictably, a number of public education groups attacked what they felt was Mr. Coleman's unjustified disparagement of public schools in favor of the private sector, and his implication that private schools may be more worthy of tax-based support than their public counterparts.
But more disturbing, if less well publicized, is the skepticism of many of Mr. Coleman's fellow social scientists, who have maintained the report is methodologically unsound and draws conclusions that are unsupported by the data.
Arthur S. Goldberger, research professor of economics at the University of Wisconsin, puts it more harshly than most, but the thrust of his criticism is typical: ''The quality of documentation, analysis, and interpretation is so defective that it is hard to avoid the overall conclusion that the report reeks with incompetence and irresponsibility."
And N.C.E.S. has subjected the report to unusually rigorous review pending its publication in final form. Even before the report was released, the agency circulated its own advance rebuttals. Later, N.C.E.S. officials asked the National Academy of Science to convene a panel of statisticians, who in July spent a day privately with Mr. Coleman combing through the report.
"It has gotten a little rough at times," Mr. Coleman said in a recent interview. And he acknowledged that he intended to make some technical revisions in his final report. But Mr. Coleman seems unperturbed by the criticism and the conflicting evidence from other studies.
"All the subsequent analyses we've done and all the other proposals have led to the same conclusions," he said. ''They are stable results which are not going to be changed by different kinds of analyses."
Some of Mr. Coleman's findings have not been in serious dispute. For example, the N.O.R.C. survey found that private high schools are more orderly and more demanding of students than are public schools. Students in private high schools reported more homework, better attendance, and higher levels of enrollment in academically rigorous courses than did their public-school counterparts.
And some of Mr. Coleman's critics believe he may be right about private schools; they just don't believe he has proven his point.
"I think his conclusions about the effectiveness [of] private high schools will be supported by further analysis," said Daniel P. Antonoplos a senior research associate at N.I.E. But Mr. Antonoplos joins others in contending that Mr. Coleman has at best shown a link between private schools and high achievement- not the cause-and-effect relationship Mr. Coleman claims.
One major unanswered question, even in Mr. Coleman's mind, is the effect of "self-selection": Are the private schools' policies themselves responsible for the students' relative success, or is it parental support for those policies that spurs students to higher achievement?
Mr. Coleman concedes the latter possibility, but says he knows of no way to account for it statistically. "We tried a number of things and none of them made much difference. That isn't to say we have solved the problem of selection bias."
Other criticisms have centered on these major issues:
- The representativeness of the sample. Only 27 non-Catholic private schools were included in the tabulations--too few, say Mr. Coleman's critics, to draw generalizations about the country's highly diverse non-Catholic private high schools as a group.
- The reliability of the students' responses, particularly on questions regarding family income and parents' educational background. Both are crucial factors in assessing socioeconomic status, and later adjustments in the study were based in part on socioeconomic status.
- The measure of segregation
used in the study. Mr. Coleman's
draft. report says that while minority
students are underrepresented
in private schools as a whole. they
are more evenly distributed among individual schools than are minority
students in public schools. Thus, the
report concludes, private schools are
less segregated than public schools.
In a later analysis, Ellis B. Page, a professor of psychology at Duke University and chairman of the national planning committee for High School and Beyond, applied a different integration standard to the Coleman data, with dramatically different results. Mr. Page measured integration as the relative ability of students of different races to meet and found the public schools twice as integrated as private schools.
- The predicted effect of a government
policy easing the financial
burden of attending a private
school. The Coleman report
predicted that, given a $1,000 annual
increase in family income, the
proportion of blacks and Hispanics
in private schools would increase.
Critics have attacked as "shaky" the econometric model leading to this conclusion. Besides, they have said, Mr. Coleman's own data show that the shift. of students from public to private schools would be negligible- less than one percent.
- Students' backgrounds. Mr.
Coleman and his associates accounted
in their study for some factors
that are known to bear on student
achievement, including parents'
educational attainment, family income,
and ethnic origin.
But N.C.E.S. analysts and other researchers believe the achievement gap would have been insignificant if the study had taken into account additional variables such as the father's occupational status, the extent of parental disciplinary control, and peer influences.
Another background factor not taken into account was the type of program in which the student was enrolled. Public-school students in vocational and "general education" programs, as well as those in academic tracks, were unfairly compared to private-school students, most of whom presumably are in college preparatory courses, Mr. Coleman's critics say.
- The manner in which the report was released. Mr. Coleman and his associates presented their draft. report at a press conference arranged by N.C.E.S. Duke University's Ellis Page, among many others, maintains the report should have been subjected to professional review before general release.
Because of Mr. Coleman's stature, Mr. Page said, other researchers will have a difficult time rebutting the report in the public mind.
Mr. Coleman now says he, too, regrets the way in which he made the report public. "1 think N.C. E.S. was as oblivious to the reverberations as we were," he says. "They wanted to air it so the issue' could be discussed."
Nonetheless, he notes, he doesn't understand why the report generated such virulent, prolonged attacks from his colleagues. "It isn't stunning or shocking," Mr. Coleman argues.
"If the report had been a report that showed that the stronger the academic demands, the higher the achievement, and the stronger the order and discipline, the higher the achievement, regardless of the students' background, nobody would have objected."
It is these attributes of high-performance schools, public and private, that Mr. Coleman prefers to discuss. He believes that the favorable conditions identified in the report simply are somewhat more likely to occur in private schools.
"The point is this," he says. "In the private sector and in the public sector, there are high-performance schools, and it behooves us to look at what they're doing."
He believes that students perform better when their parents feel they have some choice and control over the school-something he finds lacking in public education in its present form.
Mr. Coleman sees two ways of extending freedom of choice to lower=income families: open enrollment across public school district boundaries- so that a student from Chicago, for example, could attend any public school in the metropolitan area-and some form of federal assistance to families choosing private schools.
"People who do not have a choice, because of income and racial limitations, ought to have the same choices as people who have money," he said.
Vol. 01, Issue 01, Pages 7, 16